Adeno-associated viruses (AAVs) are naturally defective parvoviruses, containing linear, single-stranded DNA genomes, that are being developed as delivery systems (vectors) for gene therapies to treat diabetes and many other chronic diseases. In the absence of helper virus, AAV2, the first human type discovered, can establish a latent infection by episomal persistence of a circular double-stranded form of its DNA, in non-dividing cells. In dividing cells, AAV2 DNA persists by integration into the host chromosomes, with a strong preference for a region of human Chromosome 19, designated AAVS1. Our long-range goal is to use our knowledge of AAV2 to develop gene delivery vectors with locus-specific integration. Such vectors would greatly improve the safety of gene therapies targeting actively dividing cells, such as stem cells. AAV2 integration at AAVS1 is dependent on the Rep68 and Rep78 proteins, encoded by AAV2, which have the ability to tether AAV2 DNA to AAVS1 DNA by binding Rep binding sites (RBSs) in both DNAs, and the ability to nick and covalently attach to sites near the RBSs in each DNA. Recent work by our group and others suggests that free DNA ends are important for integration. This leads us to hypothesize that Rep68/78 could direct the integration of any linear DNA containing an RBS into AAVS1. Since the RBS is degenerate, any piece of DNA with a substantial length would be expected to have one. It is therefore not surprising that work from another group has shown that AAV2 Rep proteins can promote locus-specific integration of a linear recombinant adenovirus genome (Wang H, Lieber A: J Virol 2006, 80:11699-11709). Such hybrid gene delivery systems can overcome two major drawbacks with AAV vectors, their small payload capacity and generally low transducing efficiency. Our preliminary experiments indicate that co-infection of AAV2 with an integration-defective lentivirus vector, allows the lentivirus vector to integrate into AAVS1. We will clone and sequence integration junctions to confirm AAVS1 integration. We also plan to get a sense of the frequency of integration, using colony formation assays in selective medium.